From Interdependence to Globalization

Michael Zürn. Handbook of International Relations. Editor: Walter Carlsnaes, Thomas Risse, Beth A Simmons. Sage Publication. 2002.

The literature on interdependence in the 1970s and on globalization in the 1990s reveals remarkable similarities, of which two are especially striking. The first is that the interest in both interdependence and globalization can be seen as an expression of a ‘poorly understood but widespread feeling that the very nature of world politics is changing’ (Keohane and Nye, 2000: 104). The second is that both concepts never reached the status of a sound theory of world politics. While most users of these concepts realize that they challenge conventional theories of world politics and have pointed to, and in a sense created, new research agendas in international relations (IR), endeavors to formulate an interdependence or globalization theory of international relations have so far not succeeded.

Not least because of these commonalities, the more recent literature on globalization is confronted with questions like ‘What’s new?’ and ‘So what?’ In the United States at least, theorists of international relations seem to have adopted this point of view. As an example, the special issue ‘International Organization at Fifty: Exploration and Contestation in the Study of World Politics’ contains no articles on the effects of interdependence or, for that matter, the effects of globalization on the constitution of world politics. That globalization studies are suspected of being nothing more than a revival of interdependence literature, however, is only one way to interpret the common attributes of the interdependence literature and recent works on globalization. Another way would be to argue that interdependence theorists of earlier times–a first wave of literature was published before and after the First World War, a second wave in the 1970s–were right and it just took us some time to realize it and learn from it. When the First World War broke out, Sir Norman Angell commented: ‘No, we have not been successful. We have been merely right’ (cited after de Wilde, 1991: 92).

In this chapter I want to defend a third position, which emphasizes two differences between interdependence and globalization research. On the one hand, the notion of globalization differs from that of interdependence in that it refers to qualitatively different conditions. Whereas the notion of interdependence refers to a growing sensitivity and vulnerability between separate units, globalization refers to the merging of units. Evidence indeed suggests that the driving force of change itself has changed. Nevertheless, the causal mechanisms mentioned in connection with the driving forces and the ongoing change in world politics are quite similar in both fields. Therefore, a reassessment of those propositions made by both interdependence and globalization literature is called for. In addition, to the extent that the notion of globalization refers to much more than just interdependence between distinct units, it can come as no surprise that the propositions about change in world politics go much further in the current debate on globalization. While some of these conclusions may indeed be overstated, debates on themes such as the end of sovereignty and the end of democracy certainly indicate the need for a theory of world politics that reevaluates the notion of distinct territorial units–be they ontologically given as in realism or socially constructed as in constructivism–as theoretical buildings blocks.

The Driving Force: Different Forms of Interconnectedness and Social Spaces



In its most general form, dependence can be described as a situation in which a system is contingent upon external forces. The analysis of dependence in the social sciences focuses on activities of social groups as the external forces. Interdependence describes a situation of mutual dependence between social actors. Thus defined, interdependence relates to specific kinds of actions in specific issue areas (see Morse, 1976: 118). Based on the distinction between the types of social actors that figure as external forces, interdependence in international relations can be due to two factors. On the one hand, nation-states and national societies are dependent upon the activities of other states (state interdependence). In this sense, states have been dependent upon each other since the Westphalian system of states emerged. National security has always been dependent upon the decisions of governments in neighboring states–for instance, whether or not to wage war. On the other hand, the effects of given actions by a government may depend on societal developments that take place outside of its jurisdiction (societal interdependence). For instance, the effectiveness of a national environmental regulation may easily be undermined by increased emissions from outside that country. Moreover, the development of national economies and societies themselves cannot be understood without taking into account what happens elsewhere. State interdependence is constitutive for the Westphalian system of states. Without state interdependence there can be no society of states (Bull, 1977). Societal interdependence is not constitutive for the Westphalian state system, it is rather a (mostly unintended) side-effect of the growing interconnectedness between societies.

Social interconnectedness can lead to quite different forms of societal interdependence. While there are countless distinctions made in the literature (see Caporaso, 1978; Senghaas, 1994; de Wilde, 1991), some of these distinctions are especially important. The most consequential but also the most contested distinction is the one between ‘sensitivity interdependence,’ defined in terms of mutual effects, and ‘vulnerability interdependence,’ defined in terms of the opportunity costs of disrupting the relationship (Keohane and Nye, 1977: 12–15). Vulnerability implies more than sensitivity in that whereas vulnerability interdependence presupposes mutual effects, sensitivity interdependence implies nothing about the cost of altering the relationship. The difference as such is widely accepted and can easily be used to structure the history of thought on interdependence (Baldwin, 1980).

The concept of interdependence in IR therefore refers mainly to societal interdependence. Despite the common use of the term that is close to vulnerability interdependence, interdependence in IR is most often understood as mutual sensitivity, that means that events and actions taking place in one unit of the international system affect other units of it. Most importantly, the interdependence literature rests on a concept of social actors (most often governments) being structurally affected by the behavior of others (most often societies in other countries), but nevertheless autonomous. In this sense, each state ‘decides for itself how it will cope with its internal and external problems, including whether or not to seek assistance from others’ (see Keohane, 1993: 93). Such a notion of interdependence still implies a choice between unilateral and multilateral strategies, even if a unilateral strategy is less effective in terms of the degree to which the actor’s intentions have been fulfilled than a successful multilateral endeavor.

Globalization and Societal Denationalization

Globalization goes further than interdependence. Richard Cooper argues that ‘the internationalized economy of the 1960s was characterized by a sensitivity of economic transactions between two or more nations to economic developments within those nations’ (1986: 1). By contrast, the process of economic globalization describes a movement towards one integrated world market in which ‘buyers and sellers are in such free intercourse with one another that the prices of the same goods tend to equality easily and quickly’ (Cooper, 1986: 71). This distinction between an internationalized economy and the global integration of markets can be taken pars pro toto. When generalized to include all societal relations, it highlights the most important difference between interdependence and globalization. Globalization thus describes a process of transition towards one integrated global society and away from a cluster of merely internationalized societies. Accordingly, globalization can also be seen as a process which reduces the significance of national societies thus calling into question the distinction between domestic and foreign relations. In this view, the living conditions of people and local communities have changed through globalization; distant events of all sorts have immediate consequences not only for states but for individuals’ daily lives (Held et al., 1999: ch. 1; Hirst and Thompson, 1996: 7; Holm and Sørensen, 1995: 4–5; Rosenau, 1990: 78). This notion of globalization refers to a measurable process of social change which, in turn, may or may not have causal effects on political developments. Globalization is thus neither identical with nor does it necessarily lead to the extension of political space and governance beyond the nation-state. Nor does it necessitate the formation of a world society or transnational identities. In this respect, this societally based understanding of globalization is similar to predominant notions of interdependence. It also facilitates a distinction between different issue areas and runs counter to notions of globalization that are restricted to the economic (Altvater and Mahnkopf, 1996) or the cultural sphere (Robertson, 1992).

While the term ‘interdependence’ refers to a condition, ‘globalization’ includes the notion of a process. In this context it is helpful to contrast the terms ‘interdependence’ and ‘globalism’ (Keohane and Nye, 2000: 104). The data, however, do not justify the use of the term globalism. Globalism may only be appropriate to sketch some exceptional cases such as the financial markets and some global dangers, but it is inappropriate for most other fields. Moreover, in some areas a process towards globalism, that is, globalization as a process leading to global social spaces, does not seem to be taking place at all. Many economic indicators suggest that a gradual development towards an OECD zone is taking place, but at the same time this is accompanied by an uncoupling of this zone from other regions in the world (Goldgeier and McFaul, 1992; Singer and Wildavsky, 1993). Against this background, I prefer the term ‘societal denationalization’ (Zürn, 1992, 1995). The question that then arises is whether intensified transboundary social interactions that are already at a relatively high level signify a further decline in the importance of nationally defined borders. The condition of a society can be described as denationalized when transactions within national borders are no denser than transnational transactions. The term societal denationalization–as a process–thus has the advantage that it defines a starting point (national society) of the process but leaves the end point indeterminate. Moreover, if cases can be singled out that show a clear trend towards globalization, there is no problem in interpreting them as special instances of a more general trend towards societal denationalization. Seen thus, the transboundary pollution of the Rhine is just as much a phenomenon of societal denationalization as global warming, although only the latter is genuinely global.


What indicates a transition from interdependence to globalization or societal denationalization? The measurement of interdependence has always been contested and this also holds true for globalization. While the interconnectedness of societies is a direct measure of neither interdependence nor globalization, it has often been used as an indicator. The interconnectedness of societies can be measured by the rise of transboundary transactions relative to transactions that take place within a national territory (see Beisheim et al., 1999; Deutsch and Eckstein, 1961; Garrett, 1998b; Held et al., 1999; Hirst and Thompson, 1996; Katzenstein, 1975; Reinicke, 1998; Rosecrance and Stein, 1973; Rosecrance et al., 1977). While the increased ratio of transboundary transactions on a low level is assumed to indicate a rise in interdependence between different units, the notion of globalization refers to societal connectedness to the extent that societal borders lose importance or even dissolve, indicating the merging of formerly nationally defined units. In the words of K.W. Deutsch, societal borders dissolve when there is no more critical reduction in the frequency of social transactions (1969: 99).

The objection raised now and again by economists to this approach to measurement is that by observing these transactions little can be said about real interdependence or, for that matter, globalization. For instance, changes in flow values may be due to market volatility, i.e. changes in the attractiveness of economic locations, and perfectly integrated spaces may even be characterized by lower flow values (see Garrett, 1998a: ch. 3). For this reason, economists often propose the analysis of transaction costs and convergent prices, which they claim more closely approximate the theoretical conception of integrated spaces (Frankel, 1993; Garrett, 1998a). For instance, between 1939 and 1990 average prices for air travel dropped from around 60 cents to about 10 cents per mile; international telecommunication costs have sunk by about 8 per cent per year since the late 1960s, and the price for one computer capacity unit has plummeted by 99 per cent over the past thirty years (Zacher with Sutton, 1996: 129). To a certain extent, these developments are indeed indications for growing societal interdependence.

Nevertheless, more differentiated evidence based on the direct measurement of these transactions is necessary. First, it is by no means certain that low transaction costs are a more reliable indication of integrated social spaces than the intensification of transactions. A reduction in the price of international phone calls, for instance, tells us much less about transboundary communication than an actual increase in the number of phone calls. It is not the facilitation of communication, but communication itself that constitutes the relevant social action. Second, the argument that perfectly integrated spaces do not necessarily show evidence of an increase in transactions is theoretically correct, yet of little practical significance. There are in effect no perfect, totally stable markets, but only approximations. Real-world approximations such as national markets are indeed characterized by extremely high transaction flows. Furthermore, if transaction flows are monitored over longer periods, temporary volatilities should be negligible as random noise created by periodic political events and spasmodic competitive shifts. Third, the measurement of transaction costs is technically very problematic, especially if specific national differences are taken into consideration. As a result, when it comes to operationalization, researchers who for theoretical reasons opt for measuring transaction costs ultimately have to resort to measuring the transactions themselves. As Milner and Keohane, for example, put it: ‘An exogenous reduction in the costs of international transactions (…) can be empirically measured by the growth in the proportion of international economic flows relative to domestic ones’ (1996: 4).

Against this background, empirical studies on levels of interdependence and globalization can be summarized as follows.

• Early propositions about decreasing interdependence among highly industrialized countries (Deutsch and Eckstein, 1961) and between great powers (Waltz, 1979) have proved, at least in their generalized versions, to be wrong. While it is correct that levels of economic interdependence were lower in the 1950s and 1960s than in the decades prior to 1929, economic interdependence grew again in the industrialized world in the decades after the Second World War (Katzenstein, 1975; Rosecrance and Stein, 1973; Rosecrance et al., 1977). In the 1970s, the low levels of economic and communicative interdependence between the Superpowers increased as well (Frei and Ruloff, 1988). And as early as the late 1950s and 1960s there was a dramatic surge in both the expansion of longdistance systems and the proliferation of weapons of mass destruction. As a result, any war in which weapons of mass destruction are deployed will have an impact that by far exceeds the actual area of combat.

• Whereas with respect to some indicators such as trade quotas and the proportion of international telephone calls, the growth rates leveled out slightly in the early 1980s, the latter part of the 1980s and most of the 1990s brought a sharp increase in transborder transactions in many areas such as trade, foreign direct investments and other capital flows, but also human mobility, the volume of transborder information and communication and the exchange of cultural products (Beisheim et al., 1999: 39–320). This surge led to the use of the term globalization. In most areas, the level of interconnectedness today clearly surpasses the levels of 1914.

• Many social transactions today transcend national borders, but they are neither global nor can a general tendency towards globality be observed. In fact, new social boundaries are becoming visible at the periphery of the OECD world. This is particularly evident in the economic sphere. Transborder trade primarily takes place within the three large trade blocks of the EU/EFTA, NAFTA and ASEAN. This is followed by trade between the large trade blocks, with only a small share finally left for the rest of the world. The OECD focus is even more evident if one looks at foreign direct investments. Over 91 per cent of all foreign direct investments between 1980 and 1991 went to the OECD countries and the ten leading threshold countries (Hirst and Thompson, 1996: 67; see also Reinicke, 1998: ch. 2). Communication flows indicate a similar concentration on OECD countries. A world map showing the distribution of Internet connections is particularly illustrative. It shows that even within the OECD world there are clear gravitational centers, the borders of which, however, do not coincide with national borders. Even in the USA there are only extensive networks along the two coastlines, including parts of Canada (Beisheim et al., 1999: 65).

• Within the OECD, substantial cross-national differences in market integration remain (see Garrett, 1998b). The levels of market integration are significantly higher in smaller countries than in larger ones. Moreover, a comparison of larger economies (G7) reveals that the integration of the British and German economies into the world market is higher for most economic indicators than in other G7 countries. These national differences become even more accentuated when other fields of globalization such as culture, environment, mobility etc. are taken into account.

• The spatial scope of dense transactions varies significantly between different fields and issues. While the proportion of transborder postal deliveries, cross-border phone calls and foreign direct investments, to name but a few indicators, is still below 10 per cent in all G7 countries, the proportion of foreign trade, foreign travel, foreign e-mails and the consumption of foreign culture is often above 50 per cent (Beisheim et al., 1999: 39–320).

• A completely new development is the transboundary production of some goods and bads (as opposed to the transboundary exchange of goods), which suddenly took off in the mid-1980s. One can justifiably say that these phenomena–the Internet, international crime, global climate changes and other global environmental dangers as well as the global financial markets–represent a new departure. It is these–still relatively rare–cases to which references of deterritorialization or even de-borderization (Agnew and Duncan, 1989; Albert, 1996; Appadurai, 1996; Brock and Albert, 1995; Harvey, 1989) are most pertinent and the term globalization most justifiably applies. Multinational corporations (MNCs) are often seen as the primary agents of the transnational production of goods. Indeed, trade with intermediate products has risen sharply over the past decades. At the same time, MNCs are less footloose than usually assumed, and seem to be shaped to a remarkable extent by traits of the national economy in which the head office is located (Pauly and Reich, 1997).

In spite of these findings, the academic debate on the extent of interdependence and globalization reveals several weaknesses. First, measured against the tremendous changes in world politics that are ascribed to interdependence and globalization, systematic data on the driving forces behind these phenomena are still remarkably scarce. In particular, statistics other than those on the ratio of transborder transactions to national transactions do not exist in such detail as allows systematic comparisons across time, countries and fields. Moreover, there is a clear lack of data on other possible sources of change such as the dematerialization of the economy (see Albert et al., 1999), changes in people’s values and attitudes (see Hughes, 1999; Inglehart, 1997) and individualization processes (see Beisheim et al., 1999).

This deficit itself points to another weakness: we know little about the causes and origins of interdependence and globalization. While some, in a somewhat apolitical fashion, overemphasize the role of technology (see Wriston, 1986), others highlight the role of political decisions. Among those who focus on political decisions, three arguments deserve special attention. First, the role of the Bretton Woods institutions and embedded liberalism as the background against which globalization was able to evolve is emphasized (Ruggie, 1983). Second, the decision of the US and British governments to abolish capital controls and deregulate capital markets was identified as a particularly important move (Helleiner, 1994). Third, the notion of a transnational élite that has more or less deliberately challenged the Keynesian élites is sometimes put forward as an explanation for globalization (van der Pijl, 1998). Another, more structural strand of thinking sees economic globalization as the inherently necessary outcome of an economic organization that is permanently seeking and creating larger markets. The change from a Fordist to a post-Fordist mode of production is in this view crucial for accounting for economic globalization (Jessop, 1994). While all these studies certainly contribute to a better understanding of the causes of globalization, an all-encompassing explanation is still lacking. A thorough conceptualization of globalization as an unintended and unanticipated side-effect of the interplay of technological developments and political decisions is certainly a theoretically attractive, but little studied option. In addition, different processes of globalization may have different causes: the rise of global financial markets probably needs to be explained in other ways than the ‘world risk society,’ that is, the globalization of industrially produced risks as identified by Ulrich Beck (1999). At all events, we have little systematic knowledge on the causes of globalization. In that sense Rosenau’s (1986) critique of an underdeveloped knowledge of the driving force of interdependence applies to globalization studies as well.

How do Interdependence and Globalization Give Rise to Political Change?


Structural effects can be separated into two categories. The first refers–from an outside perspective–to a reduction of distance and difference. While increased contact and the convergence of consumption patterns often lead to a heightened emphasis of difference and conflict on the side of the actors (Elkins, 1995; Holsti, 1980), the increase in transborder transactions ‘objectively’ increases transnational contacts and thus the density of transnational relations. In a material sense, people of different societies grow closer to each other and get to know each other better. Transparency between societies increases. In addition, increased transnational transactions necessarily create an economic interest in the maintenance of good transnational relations. The interest of strengthened export capital is one reason for the stability of free trade since the Second World War (Milner, 1988).

The other direct effect is best conceived as a challenge regarding the capacity of the nation-state to unilaterally reach its processual and material governance targets. Effective governance depends upon the spatial congruence of political regulations with socially integrated areas and the absence of significant externalities. As societal interconnectedness increases, national governments are increasingly confronted with difficulties in the implementation of their policies. That is to say that policies are strongly affected by externalities insofar as they are still only valid within their own national borders and do not extend over the whole socially integrated territory (Reinicke, 1998: 65). It is possible to distinguish four specific challenges to the achievement of national governance goals as a result of interdependence and societal denationalization.

As national borders no longer encompass sufficient territory to function as self-contained markets for large companies, all national legislation aimed to protect the domestic economy is challenged. Any national measure that is not harmonized at the international level separates markets and creates a barrier for the efficient development, purchase and sale of goods and services. As the barriers between different markets dissolve, R&D costs rise and product cycles grow shorter. Larger markets and unhindered cooperation with other enterprises are then deemed essential to remain competitive. In other words, in a denationalized world the ‘static efficiency costs of closure’ increase (Frieden and Rogowski, 1996: 35). If, due to tariffs, imports are more expensive in one country than in another country with a more liberal trade policy, manufacturers who need to import parts from foreign countries will be at a significant (comparative) disadvantage and will press for liberalization. In general, economic integration will increase demands for the harmonization of national policies, or for common rules that prohibit national state intervention, in order to overcome the disadvantages of political segmentation and maximize the gains from economic exchange. These demands are due to efficiency pressures and express a desire for non-discrimination in the markets.

Political regulations may have little impact if they cover only a part of the relevant social space. A national regulation implemented by Australia alone would do little to prevent rising cancer rates due to the depletion of the atmospheric ozone layer. Along the same lines, Germany–for good reasons–has more severe restrictions on the distribution of racist propaganda material than many other countries. However, if someone residing in the United States feeds such material into the Internet, authorities in Germany cannot legally prohibit, let alone effectively prevent these activities. One may label these challenges to the effectiveness of national policies externality problems.

Policies that may still be effective at the national level may become too costly if they do not apply to all social actors within an integrated social space. In particular, policies that create costs for the production of goods may turn out to be self-defeating in terms of competitiveness for the area to which the policy applies. Manufacturers’ associations throughout the industrialized world complain at every opportunity that the social and environmental costs of production are too high. In their view, it is essential to cut wages, social policies, environmental regulations and corporate taxes. Against this background, the widespread fear of a downward spiral of national social and environmental standards is not surprising. In the national context, this discourse about competitiveness benefits those groups in particular who are not in favor of cost-intensive market-correcting or redistributive policies.

Effective participation depends on the spatial congruence between the rulers (the nation-state) and the ruled (the national society). Yet this notion becomes problematic as soon as the nature of the relevant community is contested, as happens in the course of societal denationalization (see Held, 1995). The rise of cross-border transactions infringes on the normative dignity of political borders and national identities. If there is no congruence, then a group affected by a decision but not participating in its making can be considered as being subject to foreign determination rather than self-determined. This new form of foreign determination tends to be mutual (yet not symmetrical), and derives from the manifold externalities that result from many political decisions having if not unlimited, at least transboundary effects. Any national decision that leads to transboundary pollution affects people who live outside the political system responsible for the decision. The specific challenge in this case arises out of representational deficits.


The structural changes cited above have not remained unanswered. These direct effects and challenges to the effectiveness of national policies do not directly translate into a decline of national society and the nation-state. The effects and challenges are serious, yet the outcome is largely determined by the political choices made by agents. Governments and other political organizations are able to respond to these challenges in a number of different ways.

Political actors may passively await the decline in effectiveness of national schemes, partly because they lack institutional imagination and prefer to stick to the institutional status quo, and partly because they can use the pretext of international pressure to further their own domestic goals such as deregulation and de-democratization. Many employers’ associations and liberal-minded governments, for instance, pursue international free-trade policies which they can instrumentalize for the purposes of national deregulation. This response is quite different from that of groups and governments which even under pressure focus on the maintenance of traditional national regulations (see Zürn et al., 2000). At the same time, one can interpret the willingness of many governments to selectively engage in intergovernmental agreements as an attempt to gain the autonomy of the executive from society (Wolf, 2000).

A second principal response can be described as fragmentative political restructuring (Holsti, 1980). Regionalist political actors may push for decentralization or even secession from nation-states in order to be able to respond as flexibly as possible in the new denationalized environment. The rise of the Lega Norte in Northern Italy and other regionalist movements elsewhere can thus be interpreted as a fragmentative political response (Woods, 1992). Alternatively, nationalist political actors may attempt to reassert the importance of national borders, especially by pushing for protectionist policies in areas like trade, culture and migration (Zürn, 1998: ch. 9).

Governments and other political organizations may also opt for integrative political restructuring in response to interdependence and societal denationalization. The incongruence between national political regulations and denationalized areas of social transactions calls into question the capacity of the nation-state to provide the very goods that made it the dominant political institution in the first place. In this predicament, governments and other political actors may endeavor to regain control by establishing new international and transnational or even supranational regimes, networks and organizations for the coordination and harmonization of their policies, that is, they may endeavor to establish legitimate governance beyond the nation-state.

What do Interdependence and Globalization Explain?

‘What are the characteristics of world politics under conditions of extensive interdependence?’ (Keohane and Nye, 1977: 19). Having described the rise of interdependence and globalization and discussed the causal mechanism the question now arises whether or not the assumed effects really take place. In this section I shall discuss two assumed effects that are ascribed to both growing interdependence and globalization.

Peace and Cooperation

The promise of early interdependence theorists was nothing less than ‘peace.’ Sir Robert Angell (1969) and Ramsey Muir (1932) emphasized the reduction of distance and difference as the mechanism through which rising interdependence would directly, though unintentionally, change world politics (see de Wilde, 1991). Angell’s prediction–made on the eve of the First World War–that war among industrialized nations had become unlikely, has frequently been ridiculed. However, a closer look at his and Muir’s work shows that both put forward highly advanced arguments with a good sense of their limitations. Czempiel (1986) elegantly framed this kind of thinking in terms of ‘peace through trade’ as one of the three most important peace strategies at hand. Although it is contested to what extent there is any direct association between interdependence and peace, empirical studies show at least indirect effects. Accordingly, the pacifying effects of trade depend on the symmetry and the extent of the ensuing interdependence and other factors (Barbieri, 1996; Dorusen, 1999; Polachek et al., 1999; Reuveny and Kang, 1998; Russett et al., 1998).

Keohane and Nye (1977: ch. 2) have discussed related issues using a more developed framework. In their view, interdependence does not lead to the disappearance of conflicts, but rather conflicts will take on new forms. In this context the responses of political actors are more important for an understanding of the effects of interdependence than the direct structural effects. Their model of ‘complex interdependence’ implies the hypothesis that growing interconnectedness may lead to political processes that differ from those typically described by realism. Instead of an international politics being dominated by states that focus on security as their goal and the armed forces as the means, in a situation of complex interdependence actors also include international organizations. Moreover, security and the armed forces lose relevance as a goal and as an instrument of international politics. Today it can safely be stated that in those regions of the world that are very closely interconnected, complex interdependence indeed seems to comprehend processes and outcomes of world politics much better than realism. It is hard to deny that the relative importance of issues other than security, especially economic and ecological issues, has increased over the past three decades. In addition, the role and importance of the armed forces has declined in the conduct of foreign policy between those industrialized nations that are most closely connected.

Most importantly, over the past three decades the role of international institutions has been enhanced more or less parallel to the rise of interdependence. While the number of international organizations, which is only a very rough measure for the development of international governance, has remained more or less constant (see Rittberger and Zangl, 1995; Shanks et al., 1996), the overall number of multilateral treaties deposited at the United Nations shows linear growth from less than 150 in 1960 to well over 400 in 1998. The same applies to the annual ratification of multilateral treaties (Hirschi et al., 1999: 40). The increase in international agreements is accompanied by a growing intensity of transgovernmental relations. Various state agencies–regulatory agencies, courts, executives and increasingly also legislatures–now network with their counterparts abroad (Slaughter, 1997: 190). In contrast to earlier interdependence writers, the theorists of the 1960s and 1970s thus emphasized two ‘peace strategies’ induced by rising interdependence: the reduction of difference and the rise of international institutions in response to the need to manage interdependence.

The current globalization literature is remarkably tacit on the issue of international peace and security. Relevant statements are mostly of a very general nature. Whenever the challenges and the problems of globalization are emphasized, a general picture of violence-prone chaos in a globalized world is evoked (see Horsman and Marshall, 1995). In contrast, those neoliberals who triumphantly observe the retreat of the state–which, after all, was always the instigator of inter-state wars–prophesy the end of all such wars. By emphasizing the pressures that globalization puts on authoritarian states to foster liberalization, however, some writers–more implicitly than explicitly–have also connected globalization with the third peace strategy identified by Czempiel (1986): the ‘democratization’ of authoritarian societies. In this sense, globalization may be helpful in supporting three processes that are conducive to the absence of war between states. These are the direct effects of the reduction of difference, that is, a diminished role for the armed forces (peace through trade), a liberalization of society brought about by the pressure to improve efficiency (peace through democracy) and the strengthening of international institutions as a political response (peace through international organization) (Russett et al., 1998). The evolving patterns resemble what Karl Deutsch once described as the conditions and the processes leading to a pluralistic security community (Deutsch, 1957).

The association of peace with globalization must be qualified for several reasons. First, it applies only to those areas in the world in which interconnectedness is highly prevalent and in which the effect of smaller differences is accompanied by the rise of international institutions and liberal societies. Whereas the OECD world may be described as a zone of peace, this certainly cannot be said of the rest of the world, which again proves the point that the concept of globalization, if taken literally and assessed on a global scale, is misleading (Singer and Wildavsky, 1993). The degree of interdependence and societal denationalization not only varies across issue areas but also across nations and regions, and this variance is extremely important when taking account of the effects. Second, only figures in the category of ‘inter-state wars’ have clearly decreased, while intra-state or civil wars have not. While the rise of societal interdependence and new social spaces may indeed have reduced the capacity of the nation-state to mobilize people for inter-state wars, this does not imply a decline in general willingness to participate in instances of organized collective violence. A denationalized world may give rise to additional conflicts leading to complex cleavages and numerous ‘small’ wars instead of large-scale wars.

Deregulation and Convergence

The second major issue in the interdependence literature is the autonomy of nation-states. As a result of the incongruence between economic and political spaces, nation-states are less able to control markets and to meet their governance targets through national policies. The growing ineffectiveness of national policies was the major theme of Richard Cooper’s (1968) contribution on the economics of interdependence. With the integration of markets, so the argument goes, growing numbers of national policies no longer work. Empirical studies on the effect of economic interdependence however did not support this expectation. On the contrary, work on national adaptation to external economic challenges demonstrated convincingly that domestic structures are decisive for an understanding of national political responses (see Cameron, 1978; Gourevitch, 1978; Katzenstein, 1978, 1985). Quite contrary to the original hypothesis, this literature was instrumental in bringing the state back into (Anglo-Saxon) political science (Evans et al., 1985).

Nevertheless, Cooper’s analysis has experienced a revival. Most of the current literature on the effects of globalization takes up the argument and diagnoses ‘the end of the social democratic era’ (Scharpf, 1987), the ‘retreat of the state’ (Strange, 1996), the ‘misery of politics’ (Narr and Schubert, 1994), the ‘globalization trap’ (Martin and Schumann, 1997), a ‘race to the bottom’ or at least a ‘competition of the obsessed’ (Krugman, 1995), a competitive state (Hirsch, 1995), a Schumpeterian workfare state (Jessop, 1994) or a residual state (Cerny, 1996). For similar reasons, Ohmae (1993) goes as far as to predict the end of the state itself. Common to these studies is the notion that efficiency pressures, congruence problems and above all problems of competitiveness induce a rapid deterioration in the effectiveness of national regulations. As a result, the state retreats and gives way to economic and social deregulation. Two empirical trends should corroborate this argument. First, to the extent that changes in state intervention have external causes, we should see a convergence of different national policies. Second, to the extent that efficiency pressures and competitiveness problems dominate, we should see a downward movement, that is a general trend towards deregulation.

A number of important objections have so far been raised that challenge both elements of this hypothesis. First, higher levels of economic openness increase the demand for policies to buffer the less desirable effects of world market integration. In this sense, social policies and state intervention can be seen not only as cost-intensive burdens for efficient production, but also as a form of risk insurance in the face of increased economic openness (Garrett, 1998a; Rieger and Leibfried, 1997; Rodrik, 1997). Moreover, new growth theory suggests that many state interventions are still economically efficient and thus effective, even and especially in times of global competition (Barro, 1996; Krugman, 1994). Under certain circumstances, integrated markets may even trigger a race to the top (Vogel, 1995). Hence, higher levels of economic interdependence and globalization may well lead to more, rather than less state intervention. Finally, the way external challenges are mediated is still the single most important factor for understanding national policies. The mediation process is determined by a number of different, mainly institutional factors (Kitschelt et al., 1999; Vogel, 1996; Weiss, 1998). In addition, ideas determine ‘what states make of globalization’ (Clark, 1999). Convergence–be it downward or upward–can thus not be expected to take place. The evidence produced in these studies clearly demonstrates that neither the convergence nor the deregulation trend exist in a strong and dominant sense. The level of state expenditures has not fallen parallel to the rise of economic interdependence and globalization, nor can clear convergence processes be observed (see Bernauer, 2000 for an excellent overview). In addition, a broad-scale redistribution of state expenditures from the welfare sector to the security and R&D sectors, as predicted by the notion of a competitive state, cannot be observed (Zürn, 1998: 153–7).

The debate and evidence produced so far, however, are still inconclusive for a number of reasons.

• If it is at all possible to identify a threshold that marks the step from interdependence to globalization, it is the late 1980s and early 1990s. The data of most studies utilizes statistics on state expenditures and social expenditures and, for practical reasons, now end in the early 1990s. This is still very early, especially since quite recent data seem to show first signs of a change in trends (Kittel et al., 2000). Moreover, the importance of institutional inertia is emphasized by the data problem. For instance, although Sweden made extensive cuts in social benefits, expenditures initially grew faster than in states offering less generous benefits, because the sudden rise in unemployment activated the welfare state. In such a case, figures on state expenditure indicate neither convergence nor deregulation, in spite of a significant development in this direction.

• The absence of convergence processes cannot be equated with an absence of constraints. Case studies on current political processes in many welfare states show that constraints imposed on social policies are strongly felt and translate into uneasy compromises (Seeleib-Kaiser, 2000). While the level of unemployment expenditure did grow, the amount of money received by the individual beneficiaries dropped in almost all G7 countries.

• Studies focusing on specific policy areas could easily demonstrate a strong convergent trend towards deregulation. Deregulation in the postal and telecommunication services is a strong case in point (Grande and Schneider, 1991; Vogel, 1996). It thus seems necessary to focus much more on issue area-specific differences in order to understand the dynamics triggered by globalization (see Bernauer, 2000: ch. 8 and Scharpf, 1999: ch. 3).

• It is conceivable that challenges to the effectiveness of national policies exist and that, the absence of any political response, would lead to a downward convergence. In order to maintain or even step up a given level of intervention, states may cooperate and build common policy-making institutions. Some studies on environmental regulations exemplify this process (see Héritier et al., 1996; Jänicke, 1998; Vogel, 1995); however, most studies fail to take the international response into account.

On the aggregate level, the growth in economic interdependence in the 1970s did not lead to a convergence of national policies. In fact, quite the contrary, states proved to be relatively powerful in the face of interdependence problems. Economic and societal denationalization in the late 1980s and 1990s certainly increased efficiency pressures and accentuated the problems of congruence and competitiveness. So far, however, there is little indication that structural constraints led to generalized patterns of downward convergence. In order to have a better understanding of the challenges to the effectiveness of national policies, it seems necessary to re-focus studies on at least two counts.

First, the mode of most of the studies referred to here is mainly macro-correlative. Propositions about political changes are tested by establishing the extent to which a rise in social interconnectedness has led to new political outcomes. By contrast, the study of the way in which interdependence and societal denationalization affect political processes has been neglected. An examination of this aspect of the politics of interdependence could be extremely useful for a more accurate understanding of and predictions about policy changes, as it is rather unlikely that denationalization directly affects policies. Indeed, it seems much more plausible to assume that policy outcomes are mediated by political institutions, political actors and political struggles (Cowles et al., 2001). A much better understanding of the effects of the process can be gained by looking at ‘the politics of denationalization’ rather than policy changes, and by examining processes instead of outcomes (see Zürn et al., 2000).

Second, most studies are implicitly built on the assumption that interdependence and/or societal denationalizaton give rise to similar political processes independent of the issue under investigation, although there are in fact no a priori reasons for doing so. On the one hand, different states have, for institutional reasons, different strengths and weaknesses in adapting to external pressures. In a slightly different context, Krasner noted many years ago: ‘One state may be able to alter the structure of its medical system but be unable to construct an efficient transportation network, while another can deal relatively easily with getting its citizens around but cannot get their illnesses cured’ (1978: 58). On the other hand, the opportunity for international regulation in response to increasingly ineffective national policy measures depends on a number of conditions that vary from one issue to another, that is, according to the type of challenge. Political dynamics initiated by globalization may vary greatly depending on whether the problems we are dealing with concern efficiency, competition or congruence.

Research that focuses on peace and cooperation or on convergence and deregulation as functions of increasing societal interdependence never fundamentally challenged the theory of international relations. It essentially builds on given assumptions, for instance that national societies are separable units and state executives are agents who act rationally in the name of their principals, as well as some basic economic theorems. The same applies to most of the research that falls under the ‘globalization’ heading and aims at establishing the impact of accelerated world market integration on national regulation. These studies also use existing theoretical assumptions and theorems in order to assess the effects of interdependence on national policies and world politics. In this sense, they were never intended to culminate in an interdependence theory (writ large) of world politics. They did, however, have a theoretical impact in that they constituted a serious challenge to realism on a number of counts. Since the emergence of interdependence, research in international politics can no longer be reduced to the study of security and military issues, of peace and war. World politics today is much more than that. Moreover, interdependence research brought non-state actors and, above all, international institutions to the fore. In this sense, interdependence research can be seen as the forerunner of New Institutionalism in international relations.

The Deeper Effects of Globalization on (Global) Governance

The transition from interdependence to globalization research is not only due to different values of interconnectedness. It also indicates a significant increase in conceived effects. There is hardly a modern political institution which is not allegedly challenged, transformed or undermined by globalization. Globalization is not only said to be curbing the autonomy of nation-states and enforcing a convergence of national policies, but also disabling democracy and with it the legitimacy of national political systems, altering the nature of sovereignty and thus ultimately transforming the fundamental structures of the international system (see Beisheim and Walter, 1997; Rosenau, 1997). Globalization has intensified interdependence by the creation of new social spaces, and thus raised issues like transnational participation and transnational networks as well as the merging and interplay of political institutions that once were conceived as separable units (Keohane and Nye, 2000). In my view, the issues of democracy, transnational civil society (see Risse et al., 1999 and Chapter 13 by Thomas Risse in this volume), political fragmentation and multi-level governance are the most fascinating in this context. So far, none of these fields of debate and research has come up with conclusive results, but they do highlight interesting fields of research for the future.


Through which mechanism does globalization or societal denationalization affect democracy? A growing incongruence between social and political spaces does indeed bear upon the democratic principle according to which every individual affected by a political decision should have the chance to have a say in this decision (Held, 1995: 16; Scharpf, 1993). For a long time, the notion of a nation-state consisting of a more or less contained national society, a clearly demarcated territory and an administrative apparatus designed to provide services for that society within that territory, led political theorists to treat congruence as given. The notion of a territorially defined nation-state was hence used as a shortcut to ensure the spatial congruence between rulers (the nation-state) and subjects (the national society). Yet this notion becomes problematic as soon as the nature of the relevant community is contested, as has happened in the course of societal denationalization. The increase in cross-border transactions infringes on the normative dignity of political borders (Schmalz-Bruns, 1998: 372; see also Linklater, 1998; McGrew, 1998).

For the purposes of democracy, spatial congruence is necessary at two critical points. First, congruence between the people who are affected by a decision and their representatives in the decision-making system (input congruence) is required. If there is no input congruence, then a group affected by a decision but not participating in its making can be considered to have been determined by others rather than self-determined. All environmental externalities are cases in point. Congruence between the space for which regulations are valid and the boundaries of the relevant social transactions–that is, output incongruence–is also significant for democratic legitimacy. In a denationalized world ruled by a system of formally independent nation-states, there is a danger that political communities cannot reach a desired goal due to conditions outside their jurisdiction. For instance, a social policy desired by the majority of the population of a given political community can become unaffordable for reasons of international competitiveness. Of course, political systems have always had to take external restraints into account, but from a critical perspective it is equally vital not to simply resign and adjust normative standards to political reality.

Viewing globalization as the underlying problem of democracy challenges the still somewhat dominant view that the problem is actually constituted by moving decisions to the international level. At present–and this is where analysts are almost unanimous–this legitimacy is clearly inadequate. There are two strands of thought among those who identify a democratic deficit in the way international institutions work. One strand points to institutional deficits of international institutions that can be adjusted through reforms, provided there is the right political will. Another strand considers these suggestions as naïve and even questions the mere possibility of democratic processes beyond the nation-state because the EU and other international institutions cannot meet the social prerequisites for democracy. According to skeptics, democratic legitimacy is only possible within the framework of a demos, that is, a political community with the potential for democratic self-governance as expressed in the concept of the modern nation. Beyond the nation-state, the social prerequisite for a democratic political community–the political space–is missing. Peter Graf Kielmannsegg eloquently summarizes this point of view with respect to Europe: ‘Collective identities develop, become stable and are passed into tradition in communities of communication, of experiences and of memory. Europe, even within the narrower scope of Western Europe, has no communication community, hardly any common memory and only limited common experiences’ (1994: 27; translation by M.Z.). Hence, the connection between nation and democracy is not a historical coincidence but systematic and indissoluble. A demos as exemplified in the modern nation-state requires cultural homogeneity, and without a demos there is no democracy.

To the extent that for pragmatic reasons skeptics accept the need for decisions through international institutions, globalization leads to the uncomfortable choice between ‘effective problem-solving through international institutions’ or ‘democractic political processes.’ (see Dahl, 1994). This is not, however, a particularly convincing theoretical perspective (Grande, 2000; Zürn, 2000). In democratic terms, international institutions are a sensible response to the problems facing democracy in times of societal denationalization, as they help to redress the incongruence between social and political spaces. Theoretically, the rise of international institutions in a transnational setting, that is, the ‘emergence of denationalized governance structures’ (Joerges, 1996), helps to bring all those who are affected by a political decision into the decision-making system, thus observing the principle of ‘no taxation without representation.’ What is more, international institutions help to increase the factual freedom of political communities. Governance beyond the nation-state can therefore improve both social welfare and democracy in the face of societal denationalization. In this sense, international institutions are not the problem, but part of the solution to the problems of modern democracy.

This debate raises a number of issues that require further normative reflection and much more empirical study to be resolved. On the empirical side there are two salient issues. First, what exactly are the socio-cultural prerequisites for the application of what kind of democratic procedures? Second, to what extent are these prerequisites systematically intrinsic to the nation, and to what extent can they be generated on a transnational scale as intended or unintended side-effects of societal denationalization and as endeavors to further the development of democratic procedures in effective international institutions? The debate on the relationship between globalization (global economy) and (national) democracy has stimulated some very promising research that, however, has not yet developed far enough for results to be conclusive (see Schmalz-Bruns, 1999; Wiener, 1998, 1999; Wolf, 2000).

Political Fragmentation

Interdependence and globalization can lead to both political integration and political fragmentation. Kalevi Holsti (1980: 23) has adequately described the approach of most interdependence theorists to this double movement: ‘Analysts have been so impressed by growing interdependence that they have ignored a simultaneous and parallel process that results in increased international fragmentation.’ On the contrary, analysts of globalization have often pointed to an inherent relationship between globalization and fragmentation (see Menzel, 1998), most eloquently expressed in the formula ‘Jihad vs. McWorld’ (Barber, 1995). Nevertheless, solid empirical work concerned with identifying the causal pathways between globalization and political fragmentation has remained scarce.

It is possible to distinguish at least three different fragmentative responses to societal denationalization. First, Robert Cox (1997) and others have identified so-called anti-systemic movements that organize against globalization and the accompanying loss in substance for national democracy. In this view, Chiapas’s resistance against the Mexican state, the strikes that paralyzed Paris and much of France in 1995, and protests that halted WTO negotiations in Seattle are part of a double movement in which the deregulation of markets produces social forces that struggle for re-regulation. In this view, current developments are somehow similar to what has been analyzed by Karl Polanyi (1944) with respect to the marketization and the regaining of political control in England after the Industrial Revolution.

Second, an ominous factor accompanying the decline of public confidence in traditional political authority in many OECD countries is a resurgence of right-wing extremism. Societal and political denationalization has led to a nationalist response, especially among young men who feel threatened by immigration, the decline of welfare policies and national political authority in general (Betz, 1994; Cox, 1997; Zürn, 1998: ch. 9). In a similar vein, anti-modernist movements have gained strength in parallel to globalization outside of the OECD world, too. Fears of a resulting clash of civilizations and a broad, cross-cultural Jihad movement are, however, unfounded (Senghaas, 1998).

Third, the most interesting and probably best studied case of political fragmentation in response to globalization is the rise of bottom-up regionalism, including all patterns of endeavors toward political decentralization from within the particular region. At least three causal mechanisms connecting globalization and bottom-up regionalism have been identified.

• Welfare regionalism: In a denationalized economy each region endeavors to increase its attractiveness for foreign direct investment and to increase the productivity of its economy through the reduction of national policy costs. Therefore, rich regions in countries with regional disparities may see an opportunity to pursue their own economic strategies without having to provide for the interests of other, poorer regions within that nation-state. This kind of welfare regionalism, the most prominent example being the Lega Norte in Northern Italy (Schmidtke, 1996), is most likely to emerge in regions which form part of the economic core, but not the political centre of the nation-state (Woods, 1992: 58).

• Functionality of regions: The growing ineffec tiveness of national intervention into markets restricts the states’ capacity for territorial management, which is crucial to the integration of peripheral regions in the national project. In addition, the advantages of small units in an integrating world market are increasingly emphasized in concepts like ‘flexible specialization’ (Piore and Sabel, 1989), and popularized in the debate on a ‘Europe of the Regions.’

• Diminishing risks: The rise of market-making international institutions diminishes the risks of regionalism, since one of the central tasks of the nation-state was the establishment of a large market, to which they granted the various regions access that they in turn did not wish to risk losing. In the past, the fear of losing markets used to be one of the main obstacles to regionalist mobilization. In the light of the high degree of economic integration within the OECD and regional economic blocs such as the EU and NAFTA, the role of the nation-state in granting accessibility to markets has diminished (Zürn and Lange, 2000).

In sum, the political consequences of globalization change the logic of some variables used to explain regionalism. It is no longer only the political actions of the nation-state that provoke regionalist responses, but also the undermining of some traditional functions of the nation-state through globalization and political internationalization. In an increasingly competitive world market, rich regions want to rid themselves of their national commitments, while at the same time the development of market-enhancing international institutions has reduced the risks of secession and even increased incentives to organize regionally in order to be eligible for supranational resources. The evolving complexity of governance beyond the nation-state in turn creates desires to emphasize cultural differences at the regional level and to represent regional interests directly, no longer via the nation-state. Studies on the regionalist movements in Québec and Scotland have produced some evidence in favor of such mechanisms (Bernauer, 2000: ch. 7; Lange, 2000).

Transformation of the Westphalian System and the Rise of Supranationalism

Societal denationalization over the past two or three decades has led to a rise in international institutions that in effect may change the constitution of world politics. At first sight, the quantitative growth of governance with government does not seem to affect the national constellation in qualitative terms. The constitutional principles of the Westphalian system of sovereign states are intergovernmental in that governments mutually acknowledge each other as governments, thus laying the foundation for international society (see Bull, 1977). It is therefore true that since the early nineteenth century–many see the Congress of Vienna subsequent to the Napoleonic Wars as the point of departure–international institutions have assisted states in meeting their own governance targets. What is new about more recent developments is not only the sheer amount of governance with governments, but also the types and objects of regulation.

Traditionally, international institutions regulated the interaction between states, be it in the field of security (for instance alliances, arms treaties etc.) or in the economy (reduction of tariffs). The purpose of those international institutions currently emerging is mainly to assist states in regulating societal actors. Most international environmental regimes exemplify this development. Their aim is to reduce external and internal threats to the security of human beings and their environment, and the ultimate target of these regulations is society. For this purpose, they rely increasingly not only on negative regulations (i.e. regulations that prohibit states from taking certain measures), but also on positive regulations (i.e. regulations that require states to take certain measures; see Corbey, 1995; Scharpf, 1996). In addition, new issues have been taken on board by these institutions. The development of the GATT Regime in the field of economics is a case in point. The early GATT removed government restrictions at the borders, that is, tariffs on manufactured goods. Over time this increased the importance of non-tariff barriers, thus inducing demands for regulations that focus on behind-the-border issues. The Tokyo Round of negotiations (1973–9) began to deal with non-tariff barriers such as antidumping measures, government subsidies, government procurement and customs and licensing procedures. The results of the Uruguay Round (1986–94) were a major step forward in this direction (see Kahler, 1995). International institutions have thus changed in character in response to societal denationalization. Increasingly, their aim is to regulate not only the actions of state actors but also those of societal actors, and not only at the border, but also behind-the-border issues. In doing so, positive regulations have gained importance relative to negative regulations. The extent and the objects of international governance no longer easily match the notion of a sovereign state in the Westphalian system.

Supranational governance even more clearly contradicts the notion of a sovereign state in the national constellation. Supranational institutions develop rules that are considered superior to national law and involve servants who have some autonomy from national governments. The demand for supranational governance increases as the density and scope of international governance grows. As international governance covers more and more issue areas, overlaps and collisions between the jurisdictions of international regulations and other international or national regulations becomes more likely. Supranational bodies are a logical response to avoid such collisions. Moreover, the more international regimes address behind-the-border issues, which are especially difficult to monitor and have significant impacts on societal actors, the more the question of credibility arises. A logical way to increase the credibility of commitments is to develop supranational bodies that monitor regulations and resolve conflicts (see Moravcsik, 1998: 73–7).

To be sure, beyond Europe–the area in which both societal and political denationalization have moved furthest–supranational institutions are still rare. However, even here there have been some notable recent developments. First of all, the section of the new WTO on the manufactured goods trade brought in new monitoring and dispute settlement procedures to deal more effectively with behind-the-border issues which to some extent resemble the supranational role of the European Court of Justice. Moreover, the establishment of the Rome Statute of the International Criminal Court as a permanent institution in 1998 also indicates a move in the direction of supranationality. This court ‘shall have the power to exercise its jurisdiction over persons for the most serious crimes of international concern ….’ (Article 1; see Finally, insofar as infringements of rights can be brought directly before independent bodies by individuals, as in the case of the ‘Civil Covenant,’ the ‘Race Discrimination Convention’ and the ‘Convention against Torture,’ one may also speak of an element of supranationality. Given these very recent developments and the effects these agreements may have on other issue areas, it is fair to conclude that the extent to which institutions with supranational elements have emerged in global politics is much greater than was expected ten years ago.

The shape of more recent inter-, trans- and supranational institutions is hardly compatible with the traditional notion of state sovereignty in the national constellation. Governments and other political organizations do not merely sit back and watch globalization and the decline in the effectiveness of unilateral policies. They respond to the challenges by setting up new institutions, and this should not be neglected in the analysis and understanding of statehood and the constitution of world politics. The national constellation, that is the convergence of resources, recognition and the realization of governance goals in one political organization–the nation-state–seems to be in a process of transformation into a post-national constellation. The nation-state is no longer the only site of authority and the normativity that accompanies it.

This transformation process itself can be separated into different stages. The first stage can be regarded most plausibly as a more or less unintended outcome of political responses to (perceived) functional demands. The steady progression of some international regimes to the point that they now deal with positive interventions into society and behind-the-border issues is part of this first stage. The same appears to hold for the need for credibility in the design of these more ambitious regimes and the development of supranational bodies to deal with collisions between different regulations. The second stage of the transformation is much more reflective. When society and political actors begin to comprehend the change, they begin to put issues of transboundary identity and transboundary ethics onto the agenda. Pressures to improve living conditions for people of another nationality and race that live in countries thousands of miles away, as well as the debate on European identity and democracy, are first signs of this reflective stage in the transformation process.

Conclusion: Sovereignty and IR Theory at Bay?

Interdependence and globalization studies so far hardly constitute a theory of world politics. In many ways, both these strands of literature speculate and hypothesize about the political effects of increasing societal interconnectedness using already existing theories. The globalization literature differs from interdependence studies in that it assumes a qualitatively higher level of interconnectedness and, more importantly, broadens the range of themes which may be affected by it.

In a very subtle sense, however, the interdependence literature has structured the theoretical debate in international relations (IR) over the past three decades. Waltz’s (1979) theory of international politics and other realist writings can be seen as a deliberate attempt to rescue IR as an independent discipline from the logic of interdependence. In realist thinking, it was possible not only to understand IR without taking domestic politics and international institutions into consideration; the notion was reconfirmed, moreover, that national societies and their respective states can be conceptualized as utterly separate entities. Thus the interaction of those entities was declared a field of the discipline of IR. This reconfirmation put interdependence writers on the defensive. Their counter-attack was directed at a different target: the issue of international institutions. In this way, they indirectly reconfirmed the strict notions of distinctly separate national societies and traditional notions of sovereignty. In a sense, the debate as a whole accepted the analytical shackles of ‘methodological nationalism.’

Methodological nationalism considers nation-states as the basic unit of all politics. It assumes that humankind is naturally distributed among a limited number of nations, which organize themselves internally as nation-states and delimit themselves externally from other nation-states (see Beck, 2000; Smith, 1979). In addition, it assumes that the external delimitation and the subsequent competition between nation-states are the most fundamental concepts of political organization. Methodological nationalism is distinct from normative nationalism, according to which each nation has the inalienable right to organize itself in its own culturally specific way. Methodological nationalism sees national self-determination as ontologically given and as the most important cleavage in the political sphere. This double premise pre-determines empirical observations, as can be seen for example in the case of aggregate statistics, which are almost exclusively categorized in national terms. It locates and restricts the political sphere to the national level.

Globalization studies that focus on issues like transnational civil society and the transformation of the Westphalian state system are now beginning to question methodological nationalism. Again, the realist response is strong and intellectually challenging, but quite defensive this time. Krasner (1999) defends the realist notion of unchanged sovereignty by arguing that it has always been wrong, that is, it has always been full of loopholes and never existed in any pure sense. Whereas this analysis is on the one hand historically rich and convincing, it downplays the effects of the institution of sovereignty on the Westphalian state system. Sovereignty has indeed never been a constraining norm, but a regulative idea like democracy or equality. The effectiveness of regulative ideas, however, cannot be rejected by pointing to internal contradictions or imperfections, as is possible when discussing constraining norms. Regulative ideas by definition cannot be flawless, and they comprise of different components that need to be evenly balanced. In the case of democracy, the mere claim to have perfected it would be considered as proof of the opposite. Krasner’s analysis fails to conceptualize sovereignty as a regulative idea, and thus overlooks the change that the content and the relative importance of sovereignty are undergoing–a change that affects the constitution of world politics in that methodological nationalism is becoming increasingly inappropriate for analysis.

This critique of methodological nationalism must not be confused with the thesis of the end of the nation-state. Nation-states will persist and will remain of central importance. Relaxing methodological nationalism rather means that the nation as the dominant organizing principle of politics can no longer be presumed, but must be given an empirical status. As the significance of governance beyond the nation-state increases, the separation of political issues into nationally defined territorial units must be conceptualized as a variable–dependent and independent–rather than a conceptual premise.

Governance beyond the nation-state extends the realm of the political beyond national borders and sovereign states. We must therefore develop a notion of a global polity which cannot only be seen as a substitute for nation-states at a higher level. All forms of governance beyond the nation-state lack a central authority or a ‘world state’ equipped with a legitimate monopoly of the use of force. Thus, governance beyond the nation-state cannot take the form of governance by government, rather, it needs to be a form of governance with governments such as we see in international institutions, or governance without governments as in transnational institutions, or in supranational governance. The interplay of different forms of governance beyond the nation-state produces polities of a new quality, as certainly attested by the European Multi-Level Governance System (see Jachtenfuchs and Kohler-Koch, 1996; Marks et al., 1996). Others also speak of an emerging global polity (see Ougaard and Higgott, 2001). Such conceptualizations do not need to rely on utopian thinking about a world state or world federalism. It seems more promising to recall the tradition of historical macro-sociology upheld by Stein Rokkan, Charles Tilly and others. As early as the 1960s and 1970s they aimed at overcoming the dominant approach of treating nation-states as logically independent cases. Instead, they advocated a more complete map of one interdependent system (Tilly, 1984: 129). It is hard to understand why these attempts to conceive of nation-states as parts of a larger polity have declined to the extent that globalization and interdependence have increased.

What are the features of such an emerging polity? What makes governance beyond the nation-state special and separates it from the framework of methodological nationalism? Contrary to many national political systems, such a post-national polity is characterized by (1) members that are corporate actors in highly organized and specialized sub-systems instead of individual citizens. The notion of individual representatives and personally influential politicians of states that we know from the analyses of national political systems loses importance. In such a system of multiple corporate actors, nation-states would then be a territorial form of interest organization within a functionally segmented system, existing side by side with functional forms of interest organization. The membership’s participation is motivated by (2) an interest in problem-solving, not so much in the political organization of a common identity. Members share the notion of upgrading the common interest rather than pursuing the common good. The organizing principle of a post-national polity is (3) the absence of a central authority that can impose rules and regulations and, therefore, the dominance of a decision-mode which is closer to unanimity than to majority rule, thus favoring bargaining and deliberation over voting as decision-making mechanisms (Eising and Kohler-Koch, 1999). The notion of a post-national polity requires the introduction of concepts such as functional differentiation and the conceptualization of world politics as a sphere in which not all actors, and not even all states have similar functions (see Buzan et al., 1993). It implies the acceptance of many forms of association located somewhere between hierarchy and anarchy, and of some elements of authority in world politics (Hurd, 1999; Rosenau, 1997). It also means conceiving world politics less as a struggle for survival in a brutish environment than as a struggle over appropriate forms of governance in a given polity (Reinicke, 1998; Zürn, 1998). Such a theory of complex governance in a denationalized world would devalue not only the borders between societies, but also the border between comparative politics and IR. In one of the most imaginative contributions in the special issue of International Organization, Helen Milner writes: ‘The degree of divergence between IR and the rest of political science has waxed and waned over the years, but in the past decade it seems to have lessened’ (1998: 760). The move from interdependence to globalization is one driving force behind this change.